Looking for Pocketsphinx wrapper for working with the CMU dict to extract phonetic transcriptions of words

[Benjamin Gorman]

I'm using pocketsphinx to generate automated transcriptions of audio files and so far I'm getting on great, and I have output such as for each word in my audio file. Just a timestamped begin/end transcription.

<key>
    <flag>if</flag>
    <time>00:00:00:090</time>
    <endtime>00:00:00:760</endtime>
    <value>1</value>
</key>

My question is I would also like the phonetic prouninciation included in the transcription. I can see there is a python wrapper for the https://github.com/cmusphinx/cmudict dictionary at https://github.com/cmusphinx/cmudict-tools

Does pocketsphinx have something like this wrapper built in somewhere and if not is there either a simple way to implement it myself or another library out there I can modify?

My current code that gets my transcription is is (with my xml file encoding excluded as it's not relevant)

while (iter != NULL)
{
    int32 sf, ef, pprob;
    float conf;

    ps_seg_frames(iter, &sf, &ef);
    pprob = ps_seg_prob(iter, NULL, NULL, NULL);
    conf = logmath_exp(ps_get_logmath(ps), pprob);

    //here is where we process the word

    new_utterance new_utt;
    new_utt.conf = conf;
    new_utt.sf = sf;
    new_utt.st = (float)sf / frame_rate;
    new_utt.ef = ef;
    new_utt.et = (float) ef / frame_rate;
    new_utt.utt = ps_seg_word(iter);

    printf("Recognised: %s %.3f %.3f %f\n", ps_seg_word(iter), new_utt.st,
           new_utt.et, new_utt.conf);

    std::string word = ps_seg_word(iter);

[Benjamin Gorman]

I came up with a solution to my problem.

My https://github.com/benjgorman/ofxAutomatedCaptions shows how I solved it for future reference.